Re: comit: ABI_MATH Luca's crash in gr_Abi_MathView

>
> * explicit unrefs not necessary if the smart pointers are used
> CVS:
> ----------------------------------------------------------------------
> CVS: Enter Log. Lines beginning with `CVS:' are removed automatically
> CVS:
> CVS: Committing in .
> CVS:
> CVS: Modified Files:
> CVS: Tag: ABIMATH
> CVS: src/af/math/xp/gr_Abi_MathManager.cpp
> CVS: src/wp/main/cocoa/bundle/Info.plist
> CVS:
> ----------------------------------------------------------------------
>
> Martin,
>
> there were some problems in the gr_Abi_MathManager.cpp file: you don't
> need to do explicit ref/unrefs if you always use smart pointers. If you
> do explicit unrefs, the risk is that you deallocate an object twice.
>
> I've patched gr_Abi_MathManager.cpp in order to fix this (the code can
> be polished another bit, but I preferred to change it as little as
> possible at this stage)
>
> Abiword still crashes on exit. According to gdb and valgrind, the
> problem seems to be
>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 16384 (LWP 16021)]
> 0x083d2ae3 in fp_MathRun::getMathManager (this=0x8bb89c8) at
> fp_MathRun.cpp:96
> 96 return getBlock()->getDocLayout()->getMathManager();
> (gdb) bt
>
> Is the MathManager properly allocated?
>

HI Luca,
        Thanks! The crash on exit I can fix. It's because we're setting
the pointer to the containing block to NULL before calling the
fp_MathRun destructor.

However I keep getting segfaults in gtkmathview whenever I try to insert
any maths.

I suspect I'm having trouble with different versions of gtkmathview being
picked up or by not configuringing gtkmathview correctly.

What is the configure string used to compile gtkmathview for AbiWord?

I've forgotten and I can't seem to find it in the mailing list archives.

Thanks Luca!

Martin

> HTH,
> --luca
>
>
>
Received on Mon Jan 17 23:33:35 2005